Meet the 2016 Data Science Summer Fellows

The UCI Data Science Initiative is very pleased to announce its 2016 Graduate Summer Fellows. The Graduate Summer Fellows program supports UCI PhD students involved in research projects that have a strong data science/data analysis component. Fellows receive a graduate research stipend of about $6,000 to develop research projects on novel and interdisciplinary topics.

The Graduate Summer Fellows program supports UCI PhD students involved in new interdisciplinary research projects that have a strong data science/data analysis component.

Graduate Student Fellows

Zachary M. LabeZachary M. Labe, Earth System Science

Making the most of Arctic sea ice thickness observations

The goal of this project is to extend and improve the Arctic sea ice thickness record by using satellite observations, modeled reanalysis, and a series of statistical methods. The results of this research and data analysis will refine our understanding of changes in Arctic sea ice in response to a warming climate.

vanessadelgadouciheadshotVanessa Delgado, Department of Sociology

An Assessment of Undocumented Student Resources and Alumni Outcomes at the University of California

In this project, we employ both quantitative web surveys and qualitative focus groups to assess the academic achievement, impact of resource provision, and employment trajectories of undocumented college students at the University of California. The findings of this study will shed light on how immigration status impacts the links between educational attainment, career development resources, and upward economic mobility. Given the rise of undocumented students at UCs each year, this research is important both theoretically and for future policy recommendations.

Chris RackauckasChris Rackauckas, Department of Mathematics

Early-Warning Signals of Metastasis in Human Breast Cancer

Using critical transition theory for fast-slow stochastic dynamical systems, our predictions generated to show that there may exist early-warning signals for breast cancer metastasis in the form of changes in probabilistic signatures. Utilizing single-cellular transcriptomic data, we wish to construct a psudo-timeseries for which we can identify predictive noise signatures.

faucettTaylor Faucett, Department of Physics and Astronomy

Reverse Engineering Physics from Neural Networks

Recent research has shown that unguided Deep Neural Networks (DNN) solving high energy physics problems can out-perform standard data analysis techniques, even with guided assistance from experts. We propose a novel machine learning architecture that allows physicists to probe the internal learning processes of a DNN, giving physicists insight into the information and relationships being uniquely captured by the network.

GulesserianSSevan Gulesserian, Department of Statistics

The Impact of Environmental Air Pollution on Pediatric Asthma Attacks: Identification of Differentially Susceptible Subpopulations

In epidemiological studies of air exposure, usually we will have one sample that makes up our dataset, which will contain the cases and control exposures. Our goal is to be able to differentiate if our sample comes from 1 or several populations in terms of baseline probability of event and exposure effect modification on event probability. 

gravel_headshot-(1).jpgJason Gravel, Department of Criminology, Law, and Society

Social and spatial constraints on co-offending relationships between gang members

This project takes advantage of a unique dataset of over 244,000 police encounters (i.e. arrests and field interviews) in Long Beach between 2008 and 2013 to create a geographically embedded social network made up of 111,714 individuals. The objective of this project is to investigate whether rivalries and gang injunctions influence criminal collaborations between gang members by comparing spatial distances between residences of members of a co-offending dyad involving gang members, non-gang members, gang members residing close to a rival gang’s territory and gang members under an injunction.  

burrellpngBurrel Vann Jr, Department of Sociology

Social Movements, Discourse, and Legalization: Marijuana Policy Since 1970

Discourse about marijuana and support for legalization have become more favorable over time. Despite this shift, little attention has been given to the influence of the marijuana movement in the process. In this project, I ask and answer two novel questions: 1) did marijuana social movement organizations, affect a shift in the discourse around marijuana? and 2) did this shift have an impact on actual support for legalization? This project will demonstrate ways in which the political consequences of social movements are mediated by discursive (cultural) change.

hinklePreston Hinkle, Department of Physics and Astronomy

Developing a new resistive pulse sensing data analysis package
Preston is developing a package of software tools and scripts that will be used to analyze data from resistive pulse sensing experiments, a technique for classifying nano- and micro-sized particles such as viruses, proteins, and cells. In collaboration with colleagues from the UCI School of Medicine, Preston will run resistive pulse experiments and use the tools developed to identify a subpopulation of cancer stem cells within a human breast cancer cell line.

Matlock_Headshot.jpgMelissa Matlock, Department of Public Health

Utilizing ArcGIS and Point Source Exposure Methodology to Spatially Analyze Valley Fever Exposure in Kern, Tulare, San Luis Obispo, and Frenso Counties in California

Valley Fever is a fungal respiratory disease that one can get by breathing in spores that live in the soil. Using mapping technology, my project will spatially analyze diagnosed cases in four counties in California.

nell_headshotColleen Nell, Ecology and Evolutionary Biology

Data in the Classroom: Interactive Web Applications Engage Biology Undergraduates in Statistics

In collaboration with Emily Abbott, Colleen is working on building interactive learning modules aimed at integrating introductory biology education and statistics for undergraduates. They are using their own dissertation research to develop Shiny apps that allow students to interact with real data and make important connections between experimental design, data visualization, and scientific interpretation. 

pluta_smDustin Pluta, Statistics

Statistical Models for High Dimensional Multi-Modal Data With Applications to Imaging Genetics

For the project, we hope to develop new statistical methods for the analysis of fMRI and genomic data to better understand associations between brain function, genetics, and behavior.  The focus will be on extending tensor regression modelling to account for genetic information, and applying these methods to a new imaging genetics data set.

Photo Not Available

Aryan Safaiea, Civil and Environmental Engineering

Characterizing high frequency temperature variability on global reef environments, with implications for resistance to bleaching

In this work, we characterize high frequency (i.e. daily to hourly) temperature variability on coral reefs using a synthesis of in situ temperature data, representative of many different global reef environments. We explore the influence of reef-scale physical and thermal environment on acclimatization to heat stress and resilience to bleaching.

Headshot_Elsberry.jpgLaura Elsberry, Ecology and Evolutionary Biology

 Biotic and abiotic factors influencing rockweed communities along California rocky shores

The goal of my research is to identify the factors (nutrient availability, nutrient uptake, temperature, species associations) that are important in determining where three species of rockweeds are found along the California coastline. The presence of rockweeds is important for community structure in the rocky intertidal because rockweeds provide valuable habitat for other seaweeds and invertebrates.